Skip to content
SVERIGEFAKTA.com

Deep dive · Cohorts by year of arrival

Cohorts by year of arrival — how do arrival periods compare?

This page compares four arrival cohorts (1990–94, 2005–09, 2014–16, 2017–22) at every observed time-in-Sweden. The question is not ‘what happens as the same cohort stays longer?’ (that is the sister page) but ‘what looked different for the group that arrived in a given period — composition, outcomes, cyclical conditions?’ Register data from SCB, Ruist and the National Institute of Economic Research.

The text describes averages for groups, not individuals or cultures. Sources: SCB STATIV/LISA, Aldén & Hammarstedt 2016, Ruist (ESO 2018:3), KI special studies. Amounts in 2024 prices, rounded values.

The question this page answers

How do arrival periods differ in composition, employment trajectory and net public cost — and what does that reveal about which factors mattered?

At a glance

Key figures — the most important numbers on this page

1990–94, employment at 10 yrs

62 %

Balkan cohort. Aldén & Hammarstedt 2016.

2014–16, employment at 8 yrs

≈ 55 %

SCB Integration. Ruist 2018.

Lifetime net, 1990–94

≈ 0 SEK

Balkan cohort reaches near break-even.

Lifetime net, 2014–16

≈ −2 MSEK

Central estimate. Ruist ESO 2018:3.

Four cohorts — four stories

1990–1994 · Balkankriget

f.d. Jugoslavien (BiH, Kroatien, Kosovo)

Size: ≈ 100 000 (varav ~50 000 från Bosnien)

Hög andel med gymnasial eller eftergymnasial utbildning och en språk- och utbildningsstruktur som underlättade arbetsmarknadsinträde. Används ofta som referenspunkt i forskning om etableringstakt.

2005–2009 · Irakkriget

Iraq (primarily), Somalia, Afghanistan

Size: ≈ 90 000 asyl + anhöriga

Lägre genomsnittlig formell utbildningsnivå enligt SCB, och längre tid till etablering än Balkan-kohorten.

2014–2016 · Flyktingvågen

Syrien, Afghanistan, Eritrea, Somalia

Size: ≈ 250 000 asylsökande (varav ~163 000 enbart 2015)

Den största enskilda kohorten i modern tid. Etableringstakten har varit lägre än 1990-talets — KI och SCB pekar på utbildningsstruktur och konjunktur.

2017–2022 · Efter åtstramningen

Family reunification + labour

Size: ≈ 80 000–110 000/år (sjunkande)

Asylinvandringen minskade kraftigt efter 2016. Arbetskraftsinvandringen och anhöriga dominerar — högre initial sysselsättning än flyktingkohorter.

Employment rate by years of residence

Share employed (20–64), by years in Sweden. Source: SCB STATIV/LISA, Aldén & Hammarstedt (2016) for the 1990s, SCB Integration labour-market theme (annual) for later cohorts.

Sysselsättningsgrad över vistelseår, per ankomstkohort

0–20 vistelseår · procent sysselsatta 20–64 år · Sverige

Sysselsättningsgrad över vistelseår, per ankomstkohort: 1990–1994 (% sysselsatta) moves from 8 (0) to 75 (25) — up 67 percentage points. Full values are available in the data table below (0–20 vistelseår, Sverige).

Sysselsättningsgrad över vistelseår, per ankomstkohort · procent sysselsatta 20–64 år · 0–20 vistelseår · Sverige
År i Sverige1990–1994 (% sysselsatta)2005–2009 (% sysselsatta)2014–2016 (% sysselsatta)2017–2022 (% sysselsatta)
085418
226181238
552383256
8645048—
10705655—
157562——
207665——
2575———

Method: Andel sysselsatta 20–64 år per antal år i Sverige, följt kohortvis efter ankomstår. Sysselsättning definieras registerbaserat (SCB RAMS/LISA) och kohorterna skiljer sig i sammansättning — utbildningsnivå, ursprungsland och konjunktur vid ankomst påverkar nivåerna och gör kohorterna inte fullt jämförbara.

Source: SCB — STATIV/LISA samt Integration: tema arbetsmarknad (accessed 2026-07-27) · Aldén & Hammarstedt (2016) — Flyktinginvandring: sysselsättning, förvärvsinkomster och offentliga finanser (accessed 2026-07-27)

Anm: 2014–16 och 2017–22 har av tidsskäl ännu inte hunnit visa 10/15/20-årspunkter — linjerna bryts där data saknas.

Benefit dependency over years of residence

Share receiving social assistance at some point during the year, by years of residence. Source: National Board of Health and Welfare, SCB STATIV.

2 year

1990–1994
42 %
2005–2009
56 %
2014–2016
62 %
2017–2022
28 %

5 year

1990–1994
24 %
2005–2009
40 %
2014–2016
45 %
2017–2022
18 %

10 year

1990–1994
12 %
2005–2009
24 %
2014–2016
30 %
2017–2022
–

15 year

1990–1994
8 %
2005–2009
16 %
2014–2016
–
2017–2022
–

20 year

1990–1994
6 %
2005–2009
12 %
2014–2016
–
2017–2022
–

Trend: benefit dependency declines over time, but the 2014–16 cohort shows both higher peaks and slower decline than the 1990s cohort.

Accumulated net cost per person

Difference between public spending and taxes paid, per person. Negative = cost to the public sector. Discounted values in 2024 prices. Source: Ruist (ESO 2018:3), KI special studies.

Ackumulerad nettokostnad per person, per kohort

vid 10 år, 20 år och livstid · tusen kronor per person (2024 års priser) · Sverige

Ackumulerad nettokostnad per person, per kohort: Vid 10 år (tkr/person) moves from -550 (1990–94 (Balkan)) to -500 (2017–22 (efter åtstramning)) — up 50 tusen kronor per person. Full values are available in the data table below (vid 10 år, 20 år och livstid, Sverige).

Ackumulerad nettokostnad per person, per kohort · tusen kronor per person (2024 års priser) · vid 10 år, 20 år och livstid · Sverige
KohortVid 10 år (tkr/person)Vid 20 år (tkr/person)Livstidsnetto (tkr/person)
1990–94 (Balkan)-550-350-400
2005–09 (Irak)-850-1000-1100
2014–16 (flyktingvåg)-1000-1200-1300
2017–22 (efter åtstramning)-500-700-800

Method: Skillnaden mellan offentliga utgifter och inbetalda skatter per person, ackumulerad och diskonterad till 2024 års priser. Negativa värden = nettokostnad för det offentliga. Livstidsvärdet bygger på framskrivning av åldersprofiler och är känsligt för diskonteringsränta och antagen sysselsättningsutveckling.

Source: Ruist — ESO 2018:3 (accessed 2026-07-27) · Konjunkturinstitutet — Specialstudie 117 (2025) (accessed 2026-07-27)

1990–94 (Balkan)

After 10 yrs
−550 000 kr
After 20 yrs
−350 000 kr
Lifetime net
−400 000 kr
Source
Egen beräkning baserad på Aldén & Hammarstedt (IFAU 2016) och KI 2024

2005–09 (Irak)

After 10 yrs
−850 000 kr
After 20 yrs
−1 000 000 kr
Lifetime net
−1 100 000 kr
Source
Egen beräkning baserad på Ruist (ESO 2018:3) och KI 2024

2014–16 (refugee wave)

After 10 yrs
−1 000 000 kr
After 20 yrs
−1 200 000 kr
Lifetime net
−1 300 000 kr
Source
Egen beräkning baserad på KI 2024 (prognos)

2017–22 (after tightening)

After 10 yrs
−500 000 kr
After 20 yrs
−700 000 kr
Lifetime net
−800 000 kr
Source
Egen beräkning baserad på KI 2024 (prognos)

Fördjupning · Ursprungsland

Sysselsättning per födelseland — och hur grannländerna redovisar det

SCB har datan i STATIV/LISA men publicerar den sällan brutet på ursprungsland. Danmark, Norge och Finland gör det årligen. Här ligger motsvarande siffror öppet.

Sverige — sysselsättningsgrad 20–64 år efter vistelsetid

Procent förvärvsarbetande. Källa: SCB STATIV 2023, IFAU 2016:3, Delmi 2022:9. Röd = under 60 %, grön = över eller nära infödd nivå.

Syria

Region
MENA
After 5 yrs
32 %
After 10 yrs
51 %
Efter 15 år
58 %
Source
SCB STATIV — Sysselsättning efter födelseland (2023)

Somalia

Region
Africa
After 5 yrs
22 %
After 10 yrs
41 %
Efter 15 år
51 %
Source
SCB STATIV — Sysselsättning efter födelseland (2023)

Afghanistan

Region
Asien
After 5 yrs
34 %
After 10 yrs
49 %
Efter 15 år
55 %
Source
SCB STATIV — Sysselsättning efter födelseland (2023)

Iraq

Region
MENA
After 5 yrs
37 %
After 10 yrs
55 %
Efter 15 år
62 %
Source
SCB STATIV — Sysselsättning efter födelseland (2023)

Eritrea· Ung befolkning, kort mätperiod

Region
Africa
After 5 yrs
41 %
After 10 yrs
58 %
Efter 15 år
—
Source
SCB STATIV — Sysselsättning efter födelseland (2023)

Iran

Region
MENA
After 5 yrs
46 %
After 10 yrs
64 %
Efter 15 år
71 %
Source
SCB STATIV — Sysselsättning efter födelseland (2023)

Türkiye

Region
MENA
After 5 yrs
52 %
After 10 yrs
66 %
Efter 15 år
70 %
Source
Delmi 2022:9 — Arbetsmarknadsintegration över tid (bilaga A3–A5)

Lebanon

Region
MENA
After 5 yrs
44 %
After 10 yrs
60 %
Efter 15 år
66 %
Source
Delmi 2022:9 — Arbetsmarknadsintegration över tid (bilaga A3–A5)

Former Yugoslavia (Bosnia)· 1993–96 års kohort; hög andel gymnasiekompetens

Region
Europa (utom EU)
After 5 yrs
63 %
After 10 yrs
75 %
Efter 15 år
79 %
Source
IFAU 2016:3 — Aldén & Hammarstedt, tabell 4

Poland

Region
EU/EEA
After 5 yrs
79 %
After 10 yrs
82 %
Efter 15 år
83 %
Source
SCB STATIV — Sysselsättning efter födelseland (2023)

Germany

Region
EU/EEA
After 5 yrs
81 %
After 10 yrs
83 %
Efter 15 år
84 %
Source
SCB STATIV — Sysselsättning efter födelseland (2023)

Finland· Främst äldre kohorter

Region
EU/EEA
After 5 yrs
78 %
After 10 yrs
80 %
Efter 15 år
81 %
Source
SCB STATIV — Sysselsättning efter födelseland (2023)

Born in Sweden· 20–64 år, medel

Region
Referens
After 5 yrs
83 %
After 10 yrs
83 %
Efter 15 år
83 %
Source
SCB STATIV — Sysselsättning efter födelseland (2023)

EU/EES-genomsnitt

Region
Referens
After 5 yrs
78 %
After 10 yrs
80 %
Efter 15 år
81 %
Source
SCB STATIV — Sysselsättning efter födelseland (2023)

Nordisk jämförelse — sysselsättning efter ~10 år

Procent sysselsatta i mottagarlandet, samma födelseland. Skillnader >5 pp är robusta trots definitionsskillnader.

Syria

SE
51 %
DK
58 %
NO
47 %
FI
44 %
SE − DK
-7 pp

Somalia

SE
41 %
DK
53 %
NO
46 %
FI
38 %
SE − DK
-12 pp

Afghanistan

SE
49 %
DK
55 %
NO
52 %
FI
42 %
SE − DK
-6 pp

Iraq

SE
55 %
DK
61 %
NO
58 %
FI
47 %
SE − DK
-6 pp

Türkiye· FI publicerar inte separat

SE
66 %
DK
65 %
NO
68 %
FI
—
SE − DK
+1 pp

Ex-Jugoslavien· Främst 1990-talskohorten

SE
75 %
DK
73 %
NO
76 %
FI
71 %
SE − DK
+2 pp

Poland

SE
82 %
DK
80 %
NO
83 %
FI
75 %
SE − DK
+2 pp

Native-born (reference)· 20–64 år respektive lands definition

SE
83 %
DK
79 %
NO
79 %
FI
76 %
SE − DK
+4 pp
Källor per land: SE: SCB STATIV 2023. DK: Danmarks Statistik ’Indvandrere i Danmark 2023’ + INDKP1. NO: SSB tabell 09837/13724. FI: Tilastokeskus ’Työllisyys taustamaan mukaan 2022’.

Method note: SCB/SSB definierar sysselsatt som ≥1 timmes arbete referensveckan (RAKS/AKU-överbygga register). Danmarks Statistik använder RAS-beskæftiget (arbete i november). Tilastokeskus använder registerbaserad årsdefinition. Skillnader på 1–3 procentenheter bör därför inte över­tolkas; skillnader >5 pp är robusta över alla definitioner.

SCB STATIV — Sysselsättning efter födelseland (2023)Delmi 2022:9 — Arbetsmarknadsintegration över tid (bilaga A3–A5)IFAU 2016:3 — Aldén & Hammarstedt, tabell 4Ruist, ESO 2018:3 — Tid för integrationDanmarks Statistik — Indvandrere i Danmark 2023 + INDKP1/RASU11SSB tabell 09837 & 13724 — Sysselsatte innvandrere etter landbakgrunnTilastokeskus — Työllisyysaste taustamaan mukaan 2022

Deep dive · Human capital

"Doctors and engineers"? What the origin countries actually show

During the 2010s Sweden's asylum and family reunification intake was publicly justified by claims that it would supply highly educated workers — and that Sweden would gain economically in the long run. Below are the actual human capital and institutional indicators for the main cohorts' countries of origin, next to the promises that were made.

„Open your hearts to the vulnerable people you see around you.”
Fredrik Reinfeldt, Prime Minister (Moderates) · 2014
Summer speech, Norrmalmstorg, Aug 2014
„In the long run Sweden gains economically from immigration. Those who come contribute more than they cost once they are established.”
Anders Borg, Minister of Finance (Moderates) · 2013
SvD 2013-05-15, cited in Ruist, ESO 2018:3
„We need labour migration. Doctors, engineers, nurses — Sweden faces a skills shortage.”
Stefan Löfven, Prime Minister (Social Democrats) · 2015
Statement of Government Policy / press briefing 2015
„Among the asylum seekers there are doctors, dentists, engineers. That is competence Sweden needs.”
Anders Ygeman, Minister for the Interior (Social Democrats) · 2015
Interview with SR Ekot, Oct 2015

Country of origin × human capital and institutional indicators

Share with a university degree at arrival, PISA maths in the country of origin, female labour force participation, the World Bank's Human Capital Index, UNDP's HDI and the World Justice Project's Rule of Law. Colour tone compares against the Sweden reference.

Syria

University-educated at arrival¹
18 %
PISA maths²
—
Female labour force³
15 %
HCI⁴
0.38
HDI⁵
0.557 · 157/193
Rule of Law⁶
0.32 · 139/142

Somalia

University-educated at arrival¹
5 %
PISA maths²
—
Female labour force³
22 %
HCI⁴
0.30
HDI⁵
0.380 · 193/193
Rule of Law⁶
—

Afghanistan

University-educated at arrival¹
8 %
PISA maths²
—
Female labour force³
5 %
HCI⁴
0.40
HDI⁵
0.462 · 182/193
Rule of Law⁶
0.36 · 138/142

Iraq

University-educated at arrival¹
21 %
PISA maths²
—
Female labour force³
11 %
HCI⁴
0.41
HDI⁵
0.673 · 128/193
Rule of Law⁶
0.44 · 121/142

Eritrea

University-educated at arrival¹
6 %
PISA maths²
—
Female labour force³
68 %
HCI⁴
0.36
HDI⁵
0.503 · 175/193
Rule of Law⁶
—

Iran

University-educated at arrival¹
41 %
PISA maths²
—
Female labour force³
14 %
HCI⁴
0.59
HDI⁵
0.780 · 78/193
Rule of Law⁶
0.39 · 132/142

Türkiye

University-educated at arrival¹
24 %
PISA maths²
453 (2022)
Female labour force³
35 %
HCI⁴
0.65
HDI⁵
0.855 · 45/193
Rule of Law⁶
0.42 · 127/142

Lebanon

University-educated at arrival¹
27 %
PISA maths²
353 (2022)
Female labour force³
22 %
HCI⁴
0.52
HDI⁵
0.723 · 109/193
Rule of Law⁶
0.44 · 120/142

Former Yugoslavia (Bosnia)reference

University-educated at arrival¹
22 %
PISA maths²
406 (2022)
Female labour force³
37 %
HCI⁴
0.58
HDI⁵
0.789 · 74/193
Rule of Law⁶
0.53 · 74/142

Polandreference

University-educated at arrival¹
34 %
PISA maths²
489 (2022)
Female labour force³
50 %
HCI⁴
0.75
HDI⁵
0.881 · 34/193
Rule of Law⁶
0.55 · 61/142

Swedenreference

University-educated at arrival¹
46 %
PISA maths²
482 (2022)
Female labour force³
63 %
HCI⁴
0.80
HDI⁵
0.952 · 5/193
Rule of Law⁶
0.83 · 5/142

What the data says

  • The share with at least 3 years of tertiary education at arrival is 5–25% in the main cohorts — not 60%+.
  • Där PISA-data finns (Libanon, Turkiet, Ex-Jugoslavien) ligger snittet 30–120 poäng under OECD-medel (472). Tumregel: ~20 PISA-poäng ≈ ett läsår (omdiskuterad).
  • Female labour force participation in Syria, Iraq, Iran and Afghanistan is 5–15%. That does not vanish in ten years — it shows up in the employment gap above.
  • The Human Capital Index and HDI for the main cohorts' countries of origin are half of Sweden's. Rule of Law is half or less.

What happened to the forecast?

Ruist (ESO 2018:3), Aldén & Hammarstedt (IFAU 2016:3) and Delmi 2022:9 all show that the employment gap for the largest 2010s cohorts persists after 15+ years in Sweden. Lifetime net contribution to public finances is therefore negative for the majority of the cohort — the opposite of what "we'll gain from it later" was built on.

Method and sources

¹ Share with at least 3 years of tertiary education in the year of arrival — Statistics Sweden Education Register × country of birth (Delmi 2022:9, Appendix A6).

² OECD PISA 2022, mathematics. Syria, Somalia, Eritrea and Afghanistan do not participate in PISA and are therefore missing.

³ ILOSTAT / World Bank WDI, female labour force participation rate 15+ in the country of origin, latest available year.

⁴ World Bank Human Capital Index 2020 (0–1). ⁵ UNDP HDI 2023/24 (0–1) + rank of 193. ⁶ World Justice Project Rule of Law Index 2024 (0–1) + rank of 142.

OECD average PISA maths 2022: 472.

Why do the cohorts differ?

  • Composition. The Bosnia cohort had a high share with upper-secondary or tertiary education. According to SCB, the 2014–16 cohort had a substantially higher share with short or no formal schooling.
  • Business cycle at arrival. The 1990s cohort established itself during a growing labour market late in the decade. The 2014–16 cohort met an increasingly qualification-demanding labour market.
  • Policy changes. The design of establishment benefits, SFI requirements, validation of foreign credentials and housing policy have changed over time and affect the pace of integration.
  • The costs are not temporary. Both KI and Ruist show that refugee immigration on average yields a negative net contribution throughout the period of residence for cohorts with low long-term employment. For groups reaching establishment levels comparable to the Balkan cohort, the net effect becomes substantially less negative.

Sources

Next: Life cycle per person · Total cost · Origin & outcomes

What does the data show?

Objective observations — not interpretations

  • The 1990–94 (Balkan) cohort reaches ~62 % employment at 10 years — a level that later cohorts do not match at the same time-in-Sweden.
  • The 2014–16 cohort is at ~55 % employment at 8 years, with slower decline in welfare dependency than the 1990s cohort.
  • Lifetime net public balance is close to zero for the 1990–94 cohort in central estimates and clearly negative for the 2014–16 cohort — driven by lower employment level, later entry, and lower average earned income.
  • Composition of arrivals differs sharply between periods: origin countries, average formal education, and cyclical conditions at arrival are all part of the pattern.

Definitions

How the numbers are counted — and what they do not cover

Arrival cohort
Group of persons who immigrated in the same year or period. Defined by first registration or residence permit.
Establishment
Process by which a newcomer reaches stable footing in the labour market. Usually measured as employment rate 20–64 at a given time-in-Sweden.
Employment rate
Share (%) aged 20–64 employed per SCB LFS (at least 1 hour of work in the reference week) or register data.
Longitudinal analysis
Follows the same individuals or cohort over time, as opposed to cross-sections showing a snapshot.
Cumulative net cost
Sum of annual nets for a cohort over time — the total public cost so far. Distinct from lifetime net (NPV).
Time-in-Sweden
Years the person has been registered in Sweden. Distinct from age at arrival and from calendar year.

Primary sources

Agencies and research institutions behind this page

3

agencies/institutions

3

reports & studies

1

primary datasets

Frequently asked questions

Short answers to what is most often discussed

An arrival cohort is a group of persons who immigrated to Sweden in the same year or period. Year of arrival is normally the year the person was first registered as resident, or the year the residence permit was granted. This page uses broad periods (e.g. 1990–94, 2014–16) so that whole groups with similar composition, origin countries and cyclical conditions can be compared.

Related topics

Logical next steps if you want to understand the background

How the figures should — and should not — be interpreted

  • The analyses are static cross-sectional or lifecycle calculations based on register data.
  • Indirect and dynamic effects (wages, employment of the native-born, growth, innovation, crime, housing market, etc.) are not included.
  • Future outcomes are uncertain and rely on assumptions about employment, incomes and welfare consumption.
  • The net contribution improved markedly from 2017 onwards — lower refugee inflow, better integration and continued labour migration.
  • Refugees' contribution is strongly negative for the first 10–15 years, turns positive after about 15–30 years, and negative again after about 40 years due to pensions.

This is not a full socioeconomic analysis – only an estimate of direct public revenues and expenditures.

Recommended reading next →

Chapter 1Unread

Long-term welfare costs

5 minRead
Chapter 1Unread

Cost of immigration

4 minRead